摘要
由于大规模社会网络中存在着多种结构,且这些结构对于大规模社会网络的研究至关重要,但现有的结构发现方法大多只能够发现单一的结构或事先确定的结构,不能较为全面的反应大规模社会网络的特征。为解决上述社会网络中多结构发现的问题,引入了混合隶属度随机块模型MMSBM(Mixed Membership Stochastic Block Model)。它不仅能够生成不同结构的网络,同时还可以根据随机等价原则发现网络中的其他结构。通过在两个不同规模的微博数据集上进行结构发现实验,结果表明MMSBM能够同时发现社会网络中的多种结构,与实际观测结构基本吻合,但其计算复杂度较高,在实际应用中仍难以推广。
There are a variety of structures existing in large‐scale social network ,the struc‐tures is vital for large scale social network research .Existing structure method can only be used to discover single structure or pre‐determined structures and will not be able to reflect the more comprehensive response characteristic of large‐scale social networks .In order to solve the problem of social network multi‐structure discovery ,this paper introduces the Mixed Membership Stochas‐tic Block Model short for MMSBM .MMSBM can generate different structures of the network , and at the same time can also be in accordance with the principle of random equivalent found these structues in the network .In this paper ,experiments on two different sizes of weibo data sets have been done ,the results show that MMSBM can discover the multi‐structure of the large scale social network ,and the discover results consistent with the observed structure .
出处
《太原理工大学学报》
CAS
北大核心
2015年第5期561-565,570,共6页
Journal of Taiyuan University of Technology
基金
国家"863"高技术研究发展计划项目基金:基于用户兴趣模型的媒体大数据内容整合与可视化技术(2014AA015204)
山西省自然科学基金项目:动态社会网络隐结构推断与演化的关键技术研究(2014011022-1)
关键词
随机块模型
社会网络
混合隶属度随机块模型
结构分析
社区
聚团
stochastic block model
social networks
mixed membership stochastic block mod-el
structural analysis
community
clique
